GLM-5.3
Language modelOpen weightsCard summarizedZ.ai's latest flagship model, delivering comprehensive advancements in complex software engineering and agent capabilities
Active (latest)Source: GLM-5.3 model card (Hugging Face) and Z.ai docs overview
Use it via API
Where to call it
- Z.ai API
glm-5.3 - Self-host (weights)
zai-org/GLM-5.3
Good for
- Complex software engineering and long-horizon coding agents
- Terminal and agent tasks (Terminal-Bench 3.0, Agents' Last Exam)
- Vulnerability discovery and code security review
Overview
GLM-5.3
Z.ai's latest flagship model, delivering comprehensive advancements in complex software engineering and agent capabilities
- Cost
- $1.40 in
- Context
- 1M
- Input / output
- $1.40 / $4.40
- Max output
- 128K
- Knows up to
- —
- Released
- 18 Aug 2026
Z.ai's flagship model, built on the same base model as GLM-5.2 with all gains from post-training. Z.ai reports a 50% gain over GLM-5.2 on its in-house Z.ai Code Bench, open-source state of the art on Terminal-Bench 3.0 and Agents' Last Exam, and a large jump in vulnerability discovery and exploitation. model card; docs overview
glm-5.3- Input / output per 1M
- $1.40 / $4.40
- Context window
- 1M
- Max output
- 128K
- Knows up to
- —
- Reasoning
- Always on (low / high / max)
- Weights
- Open · GLM-5.3 License (MIT-style; security review for Model-as-a-Service operators above $10B revenue)
About this card. A model card is the maker's own account of a model: what it's for, how it was trained and tested, and where it falls short. This page condenses Z.ai's GLM-5.3 model card (Hugging Face) and Z.ai docs overview. Every claim carries its source page.